Beruflich Dokumente
Kultur Dokumente
Contents
Executive summary 1
1. Introduction 2
About
This report has been researched and produced by the Open Data Institute, and
published in April 2018. Its lead author was Jack Hardinges. It was contributed to by
Rachel Wilson, Izy Champion, Peter Wells, Karin Christiansen, Jeni Tennison, Leigh
Dodds, Lucia Chauvet, Caley Dewhurst, Anna Scott and Tess Millar.
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 1
Executive summary
Peer-to-peer accommodation platforms enable people to make their home or spare
room available for others to rent, normally for a short period of time.
The global market for peer-to-peer accommodation has grown significantly in recent
years and is expected to be worth as much as £250bn by 2025.
In addition to this growth there is debate around the wider impacts of peer-to-peer
accommodation. For example, residents of some cities are concerned that it changes
the way people feel about their homes and communities. There is a risk that this debate
– and subsequent decisions and actions made by policymakers and other stakeholders
– will be based on assumptions rather than evidence and opinions informed by data.
The Open Data Institute (ODI) has undertaken a project to understand how data can
improve the peer-to-peer accommodation market to support businesses, consumers
and communities. As part of this work, we have conducted interviews and desk research
to better understand the stakeholders interested in the impact of peer-to-peer
accommodation and the impacts they focus on, the role of data in understanding impact
and the potential use of a data observatory.
We have made a number of findings, including that there is a diverse set of stakeholders
interested in the impacts of peer-to-peer accommodation for different purposes. We also
found that although peer-to-peer accommodation platforms hold vast amounts of data
that could be used to understand the market’s impacts, restricted access to this data
inhibits its use. We also found that data observatories could help to meet the demand for
more data and increase understanding of the impacts of peer-to-peer accommodation.
Taking these actions will help to improve the collective understanding of the impact of
the peer-to-peer accommodation market.
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 2
1. Introduction
Across a variety of sectors, platforms that enable people to exchange goods and
services are challenging traditional business models.
In finance, for example, the global peer-to-peer lending market grew by 271 per cent
to more than $130bn (£106.4bn) in 2016.1 Peer-to-peer car sharing aims to provide
an alternative to traditional vehicle ownership; market leader Drivey now has 1.5 million
users globally.2 These types of platforms and the activity or exchanges they facilitate
are often referred to collectively as the sharing or collaborative economy.
Peer-to-peer accommodation plays a significant role in this economy, and can be
broadly defined as accommodation – such as a spare room or an entire home – made
available by an existing homeowner for others to rent, normally for a short period of time.
Peer-to-peer accommodation platforms such as Airbnb, LoveHomeSwap and Wimdu
have grown rapidly in recent years. Platforms like these connect homeowners with
people looking for short-term accommodation, and compete with hotels and other
more established providers of short-term lettings.
The rapid growth of peer-to-peer accommodation has created opportunities. It is now
cheaper and easier for people to stay in many countries and cities. More people can
generate income by providing access to their spare rooms or properties. In the UK,
Airbnb has estimated that its guests and hosts have generated £3.46bn between July
2016 and July 2017.3 Studies have predicted that by 2025, the value of the peer-to-peer
sector could be as high as £250bn worldwide.4
There is ongoing debate surrounding the impacts of peer-to-peer accommodation and
the growth of the sharing economy more broadly. In addition to the clear opportunities
driven by the trend, some have voiced concerns about the negative outcomes it may
also fuel.
There is a risk that this debate will be based on assumptions rather than informed by
evidence and data. The best available evidence and data should be used to help
understand the impacts, so that we can collectively make informed decisions and take
appropriate action – such as how to encourage positive impacts or when to intervene
to mitigate negative ones.
As part of the Open Data Institute’s (ODI) data innovation programme,5 we have
undertaken a project to understand how data can improve the peer-to-peer
accommodation market to support businesses, consumers and communities.
1
Peer2Peer Finance News (2016), ‘Global P2P market worth over £106bn’,
http://www.p2pfinancenews.co.uk/2016/10/12/global-p2p-kpmg
2
Telegraph (2017), ‘Car sharing market expands as Europe's largest player Drivy parks
up in UK with London launch’,
https://www.telegraph.co.uk/business/2017/11/14/car-sharing-market-expands-europe
s-largest-player-drivy-parks
3
Airbnb (2017), Airbnb UK Insights Report,
https://www.Airbnbcitizen.com/wp-content/uploads/sites/48/2017/09/Airbnb-UK-Insight
s-Report_Final_Digital_v3.pdf
4
Financial Times (2014), ‘UK ‘sharing economy’ companies told of £9bn potential’,
https://www.ft.com/content/5e0348ac-23c3-11e4-8e29-00144feabdc0
5
Open Data Institute (2017), ‘ODI secures £6m to advance data innovation’,
https://theodi.org/article/odi-secures-6m-to-advance-data-innovation
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 3
During a discovery phase that ran between September and December 2017,6 attempts
to understand the impact of peer-to-peer accommodation stood out as a prominent
topic. The term “data observatory” was used in stakeholder interviews and workshops to
describe how data from different sources could be gathered and used to inform the
debate about this impact.
Our research consisted of structured interviews with a broad range of stakeholders,
supported by desk research. The findings of our research are described throughout
Sections 2, 3 and 4 of this report, and are summarised in Section 5. The final section
also includes a set of recommendations designed to increase the understanding of the
impacts of peer-to-peer accommodation.
6
Open Data Institute (2017), ‘Research: what are the impacts of peer-to-peer
accommodation platforms?’,
https://theodi.org/article/research-what-are-the-impacts-of-peer-to-peer-accommodatio
n-platforms
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 4
7
Open Data Institute (2017), ‘Research: what are the impacts of peer-to-peer
accommodation platforms?’,
https://theodi.org/article/research-what-are-the-impacts-of-peer-to-peer-accommodatio
n-platforms
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 5
We found that these stakeholders are interested in understanding numerous
types of impacts, described in Table 1 below:
Impact type Description
On the availability of The numbers of residential properties available to purchase or rent on a
long-term housing long-term basis. This can include comparative impacts on different types of
stock property such as houses and apartments, or by value.
On property rental The price of residential property available to rent on a long-term basis, often
prices defined as for six months or more.
On tourism and visitors The numbers of tourists or short-term visitors. This can include impacts on
lengths of stays, expenditure and types/characteristics of visitors.
On alternative forms of The performance of hotels and other forms of short-term accommodation.
short-term This can include impacts on occupancy rates, profitability and employment.
accommodation
On host income The income generated by hosts who make their properties available to
guests on a short-term basis.
On economic The contribution to economic performance, often using indicators such as
performance Gross Domestic Product (GDP). This can include impacts at local, national
and sectoral levels.
On fire, health and The prevalence and risks of fire, health and safety, and other incidents,
safety, and other including impacts on adherence to existing regulations and rules.
incidents
On emergency The use of and demand for emergency services such as the police, fire and
services rescue, and medical services.
On other public The use of and demand for public services such as transport, refuse and
services recycling, parks and non-emergency medical services.
On the desirability of The perceived attractiveness of an area to visit or move into. This includes
an area impacts on existing residents and visitors, as well as potential ones.
On the sense of The sense of community felt by residents and visitors. This can include
community in an area impacts on the way people feel about their homes, neighbours, businesses
and community.
Table 1: Types of impacts of peer-to-peer accommodation
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 6
We have mapped each of the impact types to the stakeholders whom we found
to have an interest in them in Figure 1 below:
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 7
domains: a group of national government officials with the remit of tourism policy may be
interested in the impact of peer-to-peer accommodation on levels of tourism and visitors,
on economic performance, and on alternative types of short-term accommodation.
Although different stakeholders have an interest in understanding the same types of
impact of peer-to-peer accommodation, we found that their motivations are likely to
differ. A national government official may be interested in an impact in their role as
stewards or regulators of the market;8 a platform will be interested as a market
participant. The process of developing an understanding of the impacts of peer-to-peer
accommodation and then acting on that understanding – for instance how it may be
used to inform a planning or licensing decision made by a local government official –
is not the focus of this report but is referenced at times in the following sections.
Stakeholders are also interested in the impacts of peer-to-peer accommodation at
different geographic and administrative levels. For example, a local government official
responsible for planning decisions is likely to be interested in more local impacts than
a national government official concerned with tourism across a country. Some are
interested at impacts at local, national and international scales.
The type of impacts described in Table 1 are likely to evolve and change over time.
Many of the interviewees we spoke to described the difficulty of identifying what types
of impacts to measure or attempt to understand related to peer-to-peer
accommodation, as well as the sharing or collaborative economy more broadly.
This difficulty was generally attributed to the speed at which the peer-to-peer
accommodation market has grown and its differences to traditional forms of short-term
accommodation. Despite these challenges, almost all stakeholders stated a desire to
increase their understanding of the impacts of peer-to-peer accommodation.
8
As part of this project’s discovery phase, we published a report exploring the different
types of interventions local and national authorities have made to support the
peer-to-peer accommodation market:
Open Data Institute (2018), ‘Exploring interventions to support the peer-to-peer
accommodation sector’,
https://theodi.org/article/exploring-interventions-to-support-the-peer-to-peer-accommod
ation-sector
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 8
9
Airbnb Citizen (2017), ‘A look at the impact of home sharing across the UK’,
https://www.Airbnbcitizen.com/a-look-at-the-impact-of-home-sharing-across-the-uk
10
Scottish Government (2017), ‘Scottish Expert Advisory Panel on the Collaborative
Economy’, https://consult.gov.scot/digital-communications/collaborative-economy
11
Airbnb (2017), ‘Airbnb Position Paper; The Collaborative Economy and Tourism in
Scotland’,
https://2sqy5r1jf93u30kwzc1smfqt-wpengine.netdna-ssl.com/wp-content/uploads/2018
/01/Airbnb-Submission-to-the-Scottish-Expert-Panel-on-the-Collaborative-Economy-No
vember-2017_Web.pdf
12
Airbnb / Halogen Communications (2017), ‘Press Release: Airbnb Unveils New
Proposals for Edinburgh’,
https://greens.scot/sites/default/files/Airbnb_Edinburgh_press_position_combinepdf.pdf
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 9
the Airbnb community in Berlin13 – are restricted to a small number of cities on a yearly
basis and were last updated for 2015.14
Airbnb also shares data on an ad hoc basis with other organisations for research
purposes, for example, to support a 2016 study into home sharing in London by
the Institute for Public Policy Research (IPPR)15 and research into the supply of
accommodation during the 2016 Olympics in Rio.16
13
https://www.Airbnbcitizen.com/data/
14
Airbnb has provided information about the methodology used to compile these
overviews:
Airbnb Citizen (2015), ‘Data Methodology’,
https://www.Airbnbcitizen.com/data-methodology
15
IPPR (2016), ‘Homesharing and London’s housing market’,
https://www.ippr.org/publications/homesharing-and-londons-housing-market
16
World Economic Forum (2016), ‘Understanding the Sharing Economy’,
http://www3.weforum.org/docs/WEF_Understanding_the_Sharing_Economy_report_20
16.pdf
17
Airdna, https://www.airdna.co/methodology
18
Inside Airbnb, http://insideAirbnb.com
19
Open Data Institute, https://theodi.org/topic/data-infrastructure
20
Open Data Institute (2018), ‘Using open data for public services’,
https://theodi.org/article/patterns-for-using-open-data-in-the-delivery-of-public-services
21
Airdna, https://www.airdna.co/services/datafeed
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 10
are for the ‘entire home or apartment’ that have a review in the last six months AND
booked frequently AND where the host has other listings.”22
Different types of organisations – including public authorities and academic institutions –
use data sourced from AirDNA and Inside Airbnb to analyse the impact of Airbnb activity.
For example, recent studies by the School of Urban Planning at the McGill University on
impacts in New York City23 and Alasdair Rae’s analysis of short-term lets in Edinburgh24
rely on data from AirDNA data and Inside Airbnb respectively. Studies into the impact
of peer-to-peer accommodation often combine data from these intermediaries with data
from other sources, such as demographic data published by local public authorities.
Some studies are then reported in popular media. For example, an article published by
Wired in January 201825 referenced a number of studies and analyses, all of which were
conducted using data, in part, from the two intermediaries.
Airbnb generally disputes the accuracy of data scraped by AirDNA and Inside Airbnb,
and by extension, the outputs produced using data provided by them. This inaccuracy
is often ascribed to the models used to infer listing occupancy and profitability. Also, as
acknowledged in the City of San Francisco's analysis of the short-term rental market,26
data related to other peer-to-peer accommodation platforms is not made available,
either directly by platforms or via third-party intermediaries. As a result, most analysis
focuses on Airbnb’s activities only, which may distort their conclusions about the
impacts of peer-to-peer accommodation, particularly for areas where other platforms are
commonly used.
Airbnb’s Terms of Service prohibit “use [of] any robots, spider, crawler, scraper or
other automated means or processes to access, collect data or other content from
or otherwise interact with the Airbnb Platform for any purpose”.27 AirDNA’s and Inside
Airbnb’s legal basis for scraping and using data from Airbnb, and making it available
to other users, is therefore unclear and questions the sustainability and legality of any
applications, services or analyses that use this data. For example, Inside Airbnb asserts
that the data it publishes is in the public domain, by publishing it under a CC0 waiver;
it is unlikely, however, that it has the rights to do that unless it has a separate agreement
with Airbnb.
There does not appear to be public debate about privacy issues related to the ways
in which data scraped from Airbnb is used. For example, although available listings
are displayed publicly to potential guests, the process of scraping data describing
an individual property or room and using it outside of its original purpose – such as by
enabling users to analyse patterns over time or to download and combine it with other
datasets – may increase the risk that individual hosts could be identified. Currently, many
decisions related to privacy sit with third-party intermediaries who scrape listing data –
22
Inside Airbnb, http://insideAirbnb.com/about.html
23
Urban Politics and Governance research group McGill University (2018), ‘The High
Cost of Short-Term Rentals in New York City’,
http://www.sharebetter.org/wp-content/uploads/2018/01/High-Cost-Short-Term-Rental
s.pdf
24
Alasdair Rae, University of Sheffield (2018), ‘Analysis of Short-Term Lets Data for
Edinburgh’,
https://greens.scot/sites/default/files/AnalysisShortTermLetsDataforEdinburgh_AlasdairR
ae_PDF-FINAL_2.pdf
25
Rowland Manthorpe (2018), ‘Airbnb is taking over London – and this data proves it’,
http://www.wired.co.uk/article/airbnb-growth-london-housing-data-insideairbnb
26
City and County of San Francisco Board of Supervisors (2015), ‘Policy Analysis
Report; Analysis of the impact of short-term rentals on housing’,
http://sfbos.org/sites/default/files/FileCenter/Documents/52601-BLA.ShortTermRentals.
051315.pdf
27
Airbnb, https://www.airbnb.co.uk/terms
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 11
Inside Airbnb states that the data it makes available for download “has been analysed,
cleansed and aggregated where appropriate to facilitate public discussion”.28
Who can access data is the characteristic that most affects how it can be used to
support decision-making or drive innovation. Data exists on a spectrum29 of closed,
shared or open, as shown below in Figure 2.
Figure 2: The Data Spectrum
Some of the types of data used to understand the impacts of peer-to-peer
accommodation are made available as open data that anyone can access, use and
share. This is mainly limited to data held by the public sector (described in Section 3.4),
such as demographic characteristics of different areas and the outputs of business
surveys. This data can be used without restriction in various studies and analyses.
Other data types exist within the shared part of the spectrum; this includes the data
Airbnb sometimes shares with organisations for research purposes. The choice made by
Airbnb and other peer-to-peer accommodation platforms not to publish, or share more
widely, data that they hold has led to third-party intermediaries like AirDNA and Inside
Airbnb scraping data from their websites. These intermediaries then make their own
decisions about how users can access the data, such as by restricting it to users who
pay for premium services or publishing it under certain licences.
28
Inside Airbnb, http://insideAirbnb.com/get-the-data.html
29
Open Data Institute, https://theodi.org/about-the-odi/the-data-spectrum/
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 12
Data collected by private sector transport providers could inform cities’ decisions
about their own public transport services, investment in physical infrastructure
and much more. In response to demand for data it holds, Uber has developed
Movement, an interactive mapping tool which aggregates data collected about its
users’ journeys and makes it available to help people understand traffic flows and
patterns around cities. Similarly to Airbnb, it also shares data with some
researchers. Uber’s Head of Transportation Policy and Research,
Andrew Salzberg, has recently said that the company is also “actively looking in
a bunch of different directions to be able to share data in different ways and for
different purposes”.30
More than 100 transport planning departments around the world use Strava
Metro data,31 collected from app users tracking their running and cycling
activities. The data describes aggregated cyclist and pedestrian activity such as
popular or avoided routes, peak commute times and origin/destination zones.
Strava Metro charges cities for access to data it holds. The cost varies based on
the amount of Strava activity taking place; it was reported in 2014 that one
year’s-worth of data on cyclists in the city of Portland cost Oregon’s department
of transportation $20,000.32 Strava states that it provides access to this data “to
help improve infrastructure and safety for cyclists, runners and pedestrians”.33
Paid data-sharing agreements and proprietary tools may not satisfy the demands
that all cities have for data. In 2017 New York’s Taxi and Limousine Commission
introduced rules requiring rideshare and livery companies, including Uber and Lyft,
to share detailed data about their passengers’ journeys on a regular basis.34
UK government and local authorities have legislated bus operators to open up
data after other attempts to persuade them to do so failed.35 A recent London
Assembly Transport Committee report suggested that Transport for London (TfL)
– which publishes open data that more than 13,000 developers now use to fuel
their products and services – should expect “reciprocity”,36 where organisations
who operate in London should be required to share data back to the city.
More cities are likely to consider making legislatory interventions to increase their
access to data in the transport sector and beyond. For private operators, this is
not the only reason that attempts to sell data to local and national authorities may
not be sustainable. Users of products and services are gaining more control over
data collected by private sector organisations, including through legislation like
the General Data Protection Regulation (GDPR). They may prefer that data they
contribute or is generated about them is made available to cities to help improve
the infrastructure they use, rather than being monetised. Also, data that can be
used for similar purposes may be collected by a number of other organisations,
30
Aarian Marshall (2018), ‘Dying to Know Uber's Secrets, Data-hungry Cities get
Creative’, https://www.wired.com/story/uber-lyft-data-research-driver-pay/
31
Strava, https://metro.strava.com/?_branch_match_id=500726494267820847
32
Telegraph (2014), ‘GPS big data: making cities safer for cyclists’,
https://www.telegraph.co.uk/technology/news/10818956/GPS-big-data-making-cities-s
afer-for-cyclists.html
33
Strava, https://metro.strava.com/fa
34
Aarian Marshall (2017), ‘The Secret Uber Data That Could Fix Your Commute’,
https://www.wired.com/2017/02/ubers-coughing-data-nyc-fix-commute/
35
House of Commons (2017), ‘Bus Services Act 2017’,
http://researchbriefings.parliament.uk/ResearchBriefing/Summary/CBP-7545#fullreport
36
UK Authority Data For Good (2018), ‘Transport for London “should obtain data from
apps”’,
http://www.ukauthority.com/data4good/entry/7940/transport-for-london-should-obtain-
data-from-apps
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 13
products and services and so may not have unique value. In transport, Google,
Apple and telecoms operators are likely to hold similar data to the transport
providers about passenger trips.37
Rather than defer choices related to data access to legislation, private sector
organisations could work collaboratively with cities to make the data they hold
as open as possible while respecting privacy and commercial confidentiality.
In the context of bike-sharing operators, Peter Wells has suggested that “if the
bike sharing companies don’t decide to be smart then I suspect the genuinely
‘smart cities’ will make the decision for them. Bike sharing companies will be
welcomed, but only the companies that decide to provide better services by
opening up their data. Smart companies will learn the lessons and get ahead
of that particular game.”38
Uses of data held by private sector organisations will extend far beyond those of
the cities they operate in. Regulators, citizens and other individuals may want to
use it to get involved in debates around policy. It could also support new products
and services. By making data as open as possible, private sector organisations
can support decision-making and drive innovation. Citymapper began publishing
open data in 2017 about its Smartbus routes, stop locations, schedules and
real-time arrival predictions. CityMapper was built using open data and the
organisation sees the value of releasing their own – it “encourage[s] all transport
operators to make their data freely available to enable innovation
and accessibility”.39
In some cases, there will be valid reasons why the data that cities require access
to cannot be made available as open data – for example, the risk of
re-identification of individual people whose journeys are described in the data.
There is an ongoing challenge to find the level of aggregation and access that
enables data to be as open as possible while protecting privacy and
commercial confidentiality.
As in other sectors, making valuable data more widely available will be central
to efforts to understand the impacts of peer-to-peer accommodation. As the
Scottish Expert Advisory Panel on the Collaborative Economy has recently
reported, “it is crucial that third party analysts focus on collaborating with
platforms to access data at source, rather than relying on
data-scraped information”.40
37
Peter Wells (2018), ‘Will bike sharing benefit from learning some data lessons from
other parts of transport?’,
https://hackernoon.com/bike-sharing-will-benefit-from-learning-some-data-lessons-9d7
77dbf360c
38
It is reported that Washington DC has since forced Mobike to start opening up some
data: https://twitter.com/oobr/status/973977824947253249.
39
Citymapper, https://citymapper.com/smartride/opendata
40
Scottish Expert Advisory Panel on the Collaborative Economy (2018), ‘Scottish Expert
Advisory Panel on the Collaborative Economy Report’,
http://www.gov.scot/Publications/2018/01/4152
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 14
41
VisitBritain, https://www.visitbritain.org/england-research-insights
42
STR, https://www.strglobal.com
43
Diane Coyle & Timothy Yu-Cheong Yeung (2016), ‘Understanding AirBnB in Fourteen
European cities’,
https://www.tse-fr.eu/sites/default/files/TSE/documents/ChaireJJL/PolicyPapers/2016_3
0_12_pp_understanding_Airbnb_in_14_european_cities_coyle_yeung_v.3.1.pdf
44
Iis Tussyadiah & Juho Pesonen (2015) ‘Impacts of Peer-to-Peer Accommodation Use
on Travel Patterns’,
https://www.researchgate.net/publication/282817382_Impacts_of_Peer-to-Peer_Accom
modation_Use_on_Travel_Patterns
45
Office for National Statistics (2017), ‘The feasibility of measuring the sharing economy:
November 2017 progress update’,
https://www.ons.gov.uk/economy/economicoutputandproductivity/output/articles/thefea
sibilityofmeasuringthesharingeconomy/november2017progressupdate#potential-future-s
ources-of-information
46
Philip Aldrick (2017), ‘ONS plans to include “sharing economy” transactions such as
Uber and Airbnb in GDP’,
https://www.thetimes.co.uk/article/ons-plans-to-include-sharing-economy-transactions-
such-as-uber-and-Airbnb-in-gdp-z5w5h9t2q
47
They Work For You, https://www.theyworkforyou.com/sp/?id=2017-11-08.22.0
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 15
describe – such as a report from an individual resident about the impact of peer-to-peer
accommodation on their sense of community – have a more personal focus than the
types of studies and analyses that rely on broader statistics. While seldom
representative, these stories can make abstract numbers more tangible, engaging and
useful. Access to both statistics and localised data may be needed to understand the
impacts of peer-to-peer accommodation. In these political discussions and debates,
a number of similar testimonies are sometimes used as evidence of wider trends,
or of types or severity of impact.
Data describing the impacts of peer-to-peer accommodation can also be generated by
cooperation between platforms and local public authorities.48 Our discovery research
found that some types of cooperation require property location and host identity data to
be provided to local authorities as part of a property registration process.49 Similarly, the
Mayor of London’s draft housing strategy states that the city will work with short-term
letting operators to develop an information-sharing protocol to support councils’
enforcement of the city’s 90-day rule.50 However, as data is not the focus of these types
of cooperation, it is unclear whether the data they generate is also used to understand
the impact of peer-to-peer accommodation.
48
This cooperation is sometimes mandated via interventions made by local or national
authorities, such as the introduction of regulation or other rules, and at other times is
developed in collaboration between industry and authorities, or proposed by
peer-to-peer accommodation platforms.
49
Open Data Institute (2018), ‘Exploring interventions to support the peer-to-peer
accommodation sector and the role of data’,
https://docs.google.com/document/d/1H1jBPNtX6AIPSL60iLazCmhukVG0X_05AUU3s-
a9CcQ/edit#
50
Robert Booth & Dan Newling (2016), ‘Airbnb introduces 90-day annual limit for London
hosts‘,
https://www.theguardian.com/technology/2016/dec/01/Airbnb-introduces-90-day-a-yea
r-limit-for-london-hosts
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 16
4. Data observatories
The term “data observatory” was used frequently in the original discovery interviews and
workshops we ran. It was used by different stakeholders to describe the possibility of
bringing together data to enable people to better understand the impacts of
peer-to-peer accommodation. As a result, we conducted further research into how the
term is used in other sectors and contexts.
51
Wikipedia, https://en.wikipedia.org/wiki/Local_information_systems
52
Leigh Dodds (2015), ‘What is a data portal?’,
https://blog.ldodds.com/2015/10/13/what-is-a-data-portal
53
Department for Work and Pensions (2017), ‘Data for people who don’t like data’,
https://dwpdigital.blog.gov.uk/2017/02/24/data-for-people-who-dont-like-data
54
Local Government Inform, http://lginform.local.gov.uk
55
National Network of Regional Coastal Monitoring Programmes of England,
https://www.channelcoast.org
56
East Riding of Yorkshire Council Data Observatory,
http://www2.eastriding.gov.uk/council/local-area-facts-and-figures/data-observatory
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 17
57
North Lincolnshire Data Observatory, http://nldo.northlincs.gov.uk/IAS_Live/
58
London Datastore, https://data.london.gov.uk
59
Data.gov, https://data.gov.uk/
60
European Data Portal, https://www.europeandataportal.eu
61
Wikipedia, https://en.wikipedia.org/wiki/CKAN
62
DataPress, https://datapress.com
63
InstantAtlas, https://www.instantatlas.com/extra
64
OpenDataCommunities, http://opendatacommunities.org
65
MAGiC, http://magic.gov.uk/home.htm
66
Marine Environmental Data & Information Network, http://www.oceannet.org/
67
Ethnicity Facts and Figures, Cabinet Office,
https://www.ethnicity-facts-figures.service.gov.uk
68
World Health Organisation, Global Health Observatory, http://www.who.int/gho/en/
69
Public Health England, Public Health Profiles, https://fingertips.phe.org.uk/
70
Public Health England, Network of Public Health Observatories,
http://webarchive.nationalarchives.gov.uk/20170106081009/http://www.apho.org.uk/
71
PHE Data and Analysis Tools,
https://www.gov.uk/guidance/phe-data-and-analysis-tools
72
Urban Observatory, http://www.urbanobservatory.ac.uk/
73
Web Science Trust, Web Observatories,
http://www.webscience.org/web-observatory/list-of-web-observatories/
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 18
74
Infectious Diseases Data Observatory, https://www.iddo.org/about-iddo
75
ClinicalStudyDataRequest, https://www.clinicalstudydatarequest.com/
76
Carto, https://carto.com/data-observatory/
77
Uber Movement, https://movement.uber.com/cities?lang=en-US
78
Strava, https://labs.strava.com/heatmap/#7.00/-120.90000/38.36000/hot/all
79
Strava, https://metro.strava.com/?_branch_match_id=500726494267820847
80
Imperial College London Data Observatory,
http://www.imperial.ac.uk/data-science/data-observatory/
81
Bristol is Open (2015), ‘Bristol Is Open Launches The Data Dome’,
http://www.bristolisopen.com/bristol-is-open-launches-the-data-dome/
82
Wikipedia, https://en.wikipedia.org/wiki/Project_Cybersyn
83
Kent Public Health Observatory, https://www.kpho.org.uk/
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 19
These types of data observatories share some characteristics. In most cases, they define
a particular “universe” to be observed, such as a theme or topic, or geographic area.
Most provide tools for users to explore, analyse and interpret data. Their collective
difference to data portals is that they are developed with a purpose, or set of purposes,
in mind beyond aiding in the discovery of data.
While the term “observatory” implies a certain neutrality, this purpose ties data
observatories closely towards a particular set of data users, the impacts they are trying
to understand and the decisions they are trying to make. In several cases there are
multiple observatories with a different lens on the same thing – like a city or a sector –
but meeting different sets of user needs. Data observatories might not provide
recommendations or direct action, but the choices made about the data they
incorporate, how different datasets are brought together and compared, and which
timescales are considered, facilitate some kinds of analyses and limit others.
The purpose of the data observatory, and the organisation that has made these choices,
are important factors for those using them to consider.
There are also some significant differences between the types of data observatories we
found, which highlight factors to consider when designing data observatories for different
user needs. For instance, they can allow for different granularities of data and analysis in
terms of geography and time. Some enable users to explore monthly or yearly data at
a regional level, whereas others enable users to explore data at the level of individual
real-time sensors. The data observatories also have different models for funding,
governance and sustainability – some are led by local authorities and others by
academia or industry.
84
Tarik Dogru, Makarand Mody & Courtney Suess (2017), ‘Comparing apples and
oranges? Examining the impacts of Airbnb on hotel performance in Boston’,
http://www.bu.edu/bhr/2017/06/07/Airbnb-in-boston/
85
Scottish Expert Advisory Panel on the Collaborative Economy (2018), ‘Scottish Expert
Advisory Panel on the Collaborative Economy Report’,
http://www.gov.scot/Publications/2018/01/4152
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 20
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 21
5. Findings and
recommendations
5.1. Findings
In summary of the findings described throughout this report:
Stakeholders and the impacts of peer-to-peer accommodation they focus on
1. A diverse set of stakeholders are interested in the impacts of peer-to-peer
accommodation. These stakeholders include local and national government
officials, emergency services, and hotels and alternative providers of short-term
accommodation. (p 4)
2. Stakeholders are interested in a variety of different impacts of peer-to-peer
accommodation. These range from those that are more readily quantifiable,
such as impacts on tourism and visitors, to the more abstract, including impacts
on the desirability of an area and the sense of community. (p 5)
3. Different stakeholders share interests in the same types of impacts of
peer-to-peer accommodation. For example, the impacts on tourism and visitors
are of interest to statistical organisations such as the ONS, as well as residents
and peer-to-peer accommodation platforms themselves. (p 6)
4. Stakeholders are generally interested in the impacts of peer-to-peer
accommodation for different purposes. For example, platforms will seek
to understand the impacts of their activity and the market to make decisions
to sustainably grow their businesses, while emergency services want to make
decisions about where resources should be prioritised. (p 7)
5. Stakeholders are interested in the impacts of peer-to-peer accommodation at
different geographic and administrative levels. Although some stakeholders are
clearly defined in these terms – such as both local and national government
officials – some stakeholders can be interested in impacts at multiple different
levels. For example, a peer-to-peer accommodation platform may be interested
in its impacts at local, national and international scales. (p 7)
6. We are in the early stages of understanding the types of impacts peer-to-peer
accommodation may have. The rapid growth of the peer-to-peer
accommodation market and the wider sharing or collaborative economy means
that it can be difficult to identify types of impacts to attempt to describe, provide
evidence of and understand. (p 7)
7. Stakeholders want to increase their understanding of the impacts of
peer-to-peer accommodation. Generally, the stakeholders we interviewed are
seeking to improve their understanding of the impact of the market and its
activity in order to make better decisions for their organisations. (p 7)
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 22
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 23
15. There is demand for increased access to data related to the impacts of
peer-to-peer accommodation, including via a data observatory. Developing
a data observatory may be one approach to increasing access to data.
Data infrastructure in the peer-to-peer accommodation market that is as open
as possible, whilst respecting privacy, will support a breadth of different uses
to help decision-making and drive innovation. (p15)
Data observatories
16. The term “data observatory” is used to describe a wide range of services
outside of the peer-to-peer accommodation market. Types of data observatories
we found include collections of data held by the public sector, policy-making
tools and even physical environments. (p 16)
17. Data observatories outside of the peer-to-peer accommodation market have
some similar characteristics, but also some significant differences. The majority
are defined by a theme, topic or geographic area, and provide tools for users to
work with data. There are differences, however, in the organisations responsible
for them and in the level of granularity of analysis they allow. (p 19)
18. Data observatories could help increase understanding of the impacts of
peer-to-peer accommodation. There is demand for more data that can be used
to understand the impacts of peer-to-peer accommodation. Data observatories
may help bring together datasets from different sources and provide tools for
users to explore, analyse and interpret the data. (p 19)
19. It is unlikely a single peer-to-peer accommodation data observatory will meet
all stakeholder needs. In this report we have described a diverse set of
stakeholders, impacts and types of data observatories. Given this diversity,
it may be that – alongside other types of data infrastructure – multiple data
observatories could support better understanding of the impacts of peer-to-peer
accommodation. Data observatories could be scoped by a combination of
factors, such as types of stakeholder, geography, and/or the type of impact that
needs to be understood. (p 20)
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 24
5.2. Recommendations
Improving the understanding of the impacts of peer-to-peer accommodation through
access to data, evidence and informed debate is critical to ensuring that the decisions
made by different stakeholders maximise the benefits of peer-to-peer accommodation
and mitigate its costs.
Based on the findings of our research, we have developed five recommendations
focused on improving the understanding of the impact of peer-to-peer
accommodation market:
1. Public and private sector organisations that hold data should commit to
working together to build data infrastructure that is as open as possible while
respecting privacy and commercial confidentiality.
Data infrastructure that is as open as possible will support decision-making and drive
innovation. It will make it easier for stakeholders with different needs to understand
impacts and make better decisions. Creating and using data observatories is one way
to achieve this.
Key stakeholders in creating, maintaining and improving this data infrastructure include
local and national government officials, peer-to-peer accommodation platforms and
statistical organisations.
The ODI's data infrastructure principles86 provide some guidance to help ensure that this
infrastructure is built incrementally and that it meets data user needs.
2. National government officials should facilitate and support the creation of
open data infrastructure in the peer-to-peer accommodation market and the
development of data observatories.
It is in national government's interest for local government to have the powers it needs
to make decisions that are devolved to it, and access to the data that will help it make
those decisions. National government officials should help facilitate conversations and
encourage private sector organisations to publish and share data to a useful standard.
It should also ensure that local government has any new powers that it requires to
undertake new types of intervention, and, potentially, powers to compel the release
of certain types of data from organisations if needed.
3. Local government officials should create environments that can support the
development of data observatories.
Recent ODI research has identified three patterns for using open data in the delivery
of public services.87 One of these patterns focused on using data to build services that
support more informed policy development. This pattern could describe the future
development of a data observatory, or multiple data observatories, to support better
understanding of the peer-to-peer accommodation market.
This research also identified the characteristics of environments in which these types of
services are built – such as organisational collaboration, strong technology infrastructure,
digital skills and literacy, open standards for data, and peer networks.
86
Open Data Institute (2016), ‘Principles for strengthening our data infrastructure’,
https://theodi.org/article/principles-for-strengthening-our-data-infrastructure/
87
Open Data Institute (2018), ‘Using open data for public services’,
https://theodi.org/article/patterns-for-using-open-data-in-the-delivery-of-public-services/
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 25
Creating this environment will need a senior champion – such as a local mayor, chief
executive or chief data officer – who can use their convening or regulatory powers to
bring together the multiple stakeholders in the public and private sectors who hold data
and those who want to use it to make better decisions.
4. Stakeholders with similar needs should develop data observatories
collaboratively.
Different stakeholders seeking to understand the impacts of peer-to-peer
accommodation may have shared needs that can be met by a data observatory.
These needs may cluster based on types of stakeholder (such as officials across a local
authority), geography (such as particular local area), and/or the type of impact in focus
(such as tourism).
Public sector design standards, like those developed by the UK's Government Digital
Service or the US Digital Service, provide guidance on how to design services.
Data standards are also an important part of data infrastructure – the ODI has shared
findings on approaches to developing standards and the impact of open data standards.
88
The development of data observatories may also help to build out wider data
infrastructure, as each one may highlight the need for additional types or sources
of data, which engaged data holders can work together to release, improve or share.
5. Local and national government officials should engage with different
stakeholders to inform decision-making related to the impacts of peer-to-peer
accommodation.
A better understanding of the impacts of peer-to-peer accommodation will help
policymakers to make informed decisions. It is important, however, that these decisions
are also informed by input from stakeholders such as statistical organisations, academic
and research institutions, independent researchers and residents. This input is key to
gathering data or information that is not collected, managed or used in a structured way,
such as the reports of individuals who may have benefited from the peer-to-peer
accommodation market and those who may have been adversely affected.
88
Open Data Institute (2018), ‘Open standards for data: adoption, approaches and
impact’,
https://theodi.org/article/open-standards-for-data-adoption-approaches-and-impact/
Open Data Institute 2018 / Report Impacts of peer-to-peer accommodation, the role of data and data observatories 26